- 05 Mar, 2013 40 commits
-
-
Hans Verkuil authored
Set the initial standard of subdevices instead of leaving it undefined. Also update the width and height when a new standard is chosen and return -EBUSY when attempting to change the standard while videobuf is busy. Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
There was already a core lock, so why wasn't ioctl already replaced by unlock_ioctl? This patch switches to unlocked_ioctl. Signed-off-by:
-
Hans Verkuil authored
The use of this field is deprecated since it will not work when multiple device nodes reference the same video input (the video and vbi nodes in this case). The norm field should be a device-global value. Signed-off-by:
-
Hans Verkuil authored
Required to resolve v4l2-compliance failures. Signed-off-by:
-
Hans Verkuil authored
This is needed to resolve the v4l2-compliance complaints about the control ioctls. Signed-off-by:
-
Hans Verkuil authored
Various v4l2-compliance fixes: remove unused sliced VBI functions, zero the reserved fields of struct v4l2_vbi_format and implement the missing s_fmt_vbi_cap. Signed-off-by:
-
Hans Verkuil authored
Let the tuner clamp the frequency and store that clamped value. This fixes a v4l2_compliance failure. Signed-off-by:
-
Hans Verkuil authored
The priv field of v4l2_pix_format must be zeroed. Also fix a bug in try_fmt where the current width was used instead of the width passed to try_fmt. Signed-off-by:
-
Hans Verkuil authored
The g_tuner call wasn't passed on to the subdevices, g_frequency didn't check for invalid tuners and a low-level function that was expected to return 0 or a negative error returned a positive number instead, causing s_frequency to return bogus errors. Signed-off-by:
-
Hans Verkuil authored
The audio selection code is broken. Audio and video indices were mixed up and s_audio would reject changing the audio input to something else anyway, so what's the point? All the audio input code has been removed. Signed-off-by:
-
Hans Verkuil authored
Radio should not use video or audio inputs. In addition, fix a bug in radio_g_tuner where s_tuner was called in the tuner subdev instead of g_tuner. Signed-off-by:
-
Hans Verkuil authored
This fixes a v4l2_compliance failure. [mchehab@redhat.com: CodingStyle fixes] Signed-off-by:
-
Hans Verkuil authored
This fixes a v4l2_compliance failure. Signed-off-by:
-
Hans Verkuil authored
Note that the current driver does not implement enumaudio (so apps cannot tell that audio inputs are present), it does not set V4L2_CAP_AUDIO, nor does it set audioset when calling ENUM_INPUT. And G_AUDIO doesn't set the stereo flag either. So these g/s_audio ioctls are quite pointless and misleading. Especially since some surveillance boards do not have audio at all. Signed-off-by:
-
Hans Verkuil authored
Just opening the radio tuner should not cause a switch to the radio tuner. Only after calling g/s_tuner or g/s_frequency should this happen. This prevents audio being unmuted as soon as the driver is loaded because some process opens /dev/radioX just to see what sort of node it is, which switches on the radio tuner and unmutes audio. This code can be improved further by actually keeping track of who owns the tuner and returning -EBUSY if switching tuner modes will cause problems. But for now just fix the annoying case where on boot the radio turns on automatically. Signed-off-by:
-
Hans Verkuil authored
try_fmt_vid_overlay should map incorrect sizes and fields to valid values. It also expects that an initial overlay size is defined so g_fmt_vid_overlay returns valid information. Signed-off-by:
-
Hans Verkuil authored
The capability field of v4l2_tuner should be ORed by the various subdevs and by the main driver. In this case the stereo capability was dropped. Signed-off-by:
-
Hans Verkuil authored
Probe for additional tvaudio devices, and allow tvaudio+tda7432 to co-exist. My STB TV PCI FM bttv card has a tda7432, a tda9850 and a tea6420 and with this patch it finally works again (probably for the first time in many years). Signed-off-by:
-
Hans Verkuil authored
The 'current_norm' field cannot be used if multiple device nodes (video and vbi in this case) set the same std. Signed-off-by:
-
Hans Verkuil authored
Replace the - incorrect - manual priority handling with the core priority implementation. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Note that the private chroma agc control has been replaced with the standard CHROMA_AGC control. Also fixes a mute/automute problem where closing the file handle would force mute on. That's not what you want since that would make the mute state out of sync with the mute control. Instead check against the user count. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
- don't return -EINVAL for invalid field types, handle those as if it was FIELD_ANY. - the handling of FIELD_SEQ_BT/TB was wrong as well: if such field formats aren't supported, then fall back to FIELD_ANY instead of returning an error. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Set an initial frequencies when the driver is loaded. That way G_FREQUENCY will give a frequency that corresponds with reality. Signed-off-by:
-
Hans Verkuil authored
If no tuner is present, then disable the tuner and frequency ioctls. We can remove a number of checks from those ioctls testing for the presence of a tuner. Also remove some tuner type checks (now done by the core) and fix an error return when the prio check fails. Finally some 'unlikely' statements are removed since those only make sense in tightly often executed loops, otherwise they just clutter up the code. Signed-off-by:
-
Hans Verkuil authored
- Fix ENUM_INPUT audioset. - Fix incorrect input check in s_input. Signed-off-by:
-
Hans Verkuil authored
VIDIOC_DBG_G_CHIP_IDENT is a prerequisite for the G/S_REGISTER ioctls. In addition, add support to call G/S_REGISTER for supporting i2c devices. Signed-off-by:
-
Hans Verkuil authored
The querycap ioctl didn't support V4L2_CAP_DEVICE_CAPS and the radio device implemented audio and video inputs and s_std, which are not part of the radio API. Signed-off-by:
-
Hans Verkuil authored
Remove two maintainers: telegent.com no longer exists, so those email addresses are invalid as well. Added myself as co-maintainer and change the status to 'Odd Fixes'. Signed-off-by:
-
Hans Verkuil authored
ioctl debugging can now be done through the debug parameter in sysfs. Signed-off-by:
Huang Shijie <shijie8@gmail.com> Signed-off-by:
-
Hans Verkuil authored
Due to a poor administration of the driver state it wasn't possible to open a video or vbi device multiple times. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
The usual set of problems: the frequency isn't clamped to the frequency range, no tuner index check and the frequency isn't initialized properly on module load. Signed-off-by:
-
